Skip to main content
SearchLoginLogin or Signup

The Neighborhood Context of Perceived and Reported Anti-White Hate Crimes

Hate crimes have received substantial scholarly attention, largely focusing on victims from marginal groups. Large numbers of White Americans also report being the victim of racial hate crime, though very little research has sought to examine the etiology or meaning of anti-...

Published onFeb 28, 2019
The Neighborhood Context of Perceived and Reported Anti-White Hate Crimes
·

Abstract

Hate crimes have received substantial scholarly attention, largely focusing on victims from marginal groups. Large numbers of white Americans also report being the victim of racial hate crime, though very little research has sought to examine the etiology or meaning of anti-white hate crimes.  The present work explores the neighborhood context of hate crimes against non-Hispanic whites in a majority-white city—comparing police reports with self-reported victimizations.  Police reports of anti-white hate crimes are most common in areas that have high rates of non-hate crimes and residential instability.  Perceptions of bias incidents, by contrast, appear largely driven by the racial composition.  Hate crimes against members of dominant groups appear fundamentally distinct from hate crimes against members of subordinate groups and require separate theoretical models of their substantive meaning and etiology.  In general, white residents appear to interpret the motivations for victimizations through a racial lens—attributing anti-white motivations most often when they live near larger numbers of black neighbors—and reporting them most frequently in disorganized and higher-crime places. Implications for theory and research on hate crimes, racial threat, anti-white racism, and the effect of the racial composition on perceptions of crime are discussed. 

Key Words: Hate crimes, anti-white bias, neighborhood context, racial composition, social disorganization


Hate crime refers to unlawful, violent, destructive or threatening behavior in which the perpetrator is motivated in whole or in part by prejudice towards the victim’s perceived race, ethnicity, religion, sexual orientation, gender identity, age, or impairment (Federal Bureau of Investigation 2013; Green, McFalls, and Smith 2001; Levin and McDevitt 2002).  Most journalistic accounts of hate crime, hate crime legislation, and academic research on hate crimes focus on crimes committed against members of historically oppressed or marginalized groups.  For instance, the term “hate crimes” first came into widespread use among journalists in the 1980s to describe a series of incidents targeting black, Jewish, and Asian-Americans (e.g. Levin and McDevitt 1993; Shively 2005).  The first official hate crime legislation at the state and eventually federal-level was also passed around this time (Shively 2005), though many point to precedents in the anti-KKK Civil Rights Act of 1871 and the protections from race-based injury or intimidation in the 1968 Civil Rights Act. 

Despite this focus on victims from historically oppressed groups, large numbers of white citizens also report being the victim of racially-motivated hate crimes (e.g. Lyons 2007, Gladfelter et al. 2017). In fact, as described below, in the majority-white city of Seattle, white residents report being the victim of hate crimes at lower rates but in greater numbers than non-white residents. 

Although only a small amount of research has directly considered hate crimes against members of dominant groups, the results suggest these crimes may be etiologically distinct from those perpetrated against members of subordinate groups.  Hate crimes, like non-hate crimes, cluster geographically in spaces with particular kinds of social and demographic characteristics (e.g. Green, Strolovitch and Wong 1998).  However, anti-white and anti-minority crimes appear to cluster in different kinds of places.  Two studies—Lyons’s (2007) study of neighborhoods in Chicago and Gladfelter et al.’s (2017) study of municipalities in Pennsylvania—each come to a similar but striking conclusion: unlike anti-minority hate crimes, which tend to occur in more advantaged and organized places, anti-white hate crimes appear most likely in classically socially disorganized communities.   

This research, however, relies—as does much hate crime research—on public records of hate crimes: those crimes which were reported to the police (e.g. Lyons 2007) or to the media or hate crime advocacy groups (e.g. Gladfelter et al. 2017).  As with the counting of other kinds of crimes, these public records are likely to substantially undercount hate crimes both generally and in systematic ways (Levin 1999; Iganski, 2001; Perry 2001; Bureau of Justice Statistics 2014).  In particular, public records may best capture those incidents which are both serious in their consequences and obviously motivated by bias—and especially the “stereotypical” white-on-black hate crimes (Lyons 2008b) which are most likely to be cleared and result in arrests (Lyons and Roberts 2014; Lantz et al. 2017). 

The most basic issue in identifying hate crimes is the difficulty of establishing a motive with any certainty.  This may be a particular problem for anti-white hate crimes.  While anti-minority hate crimes appear to be specialist crimes—occurring in places where other crimes are less common—anti-white hate crimes occur in the same places as other types of crimes (e.g. Lyons 2007).  In short, people may perceive an incident as being motivated by bias, but, lacking clear evidence of the bias, may not report the incident to the police as a bias crime.  Importantly, these perceptions may not always be correct.  In cases where the motive is not clearly communicated by the offender, victims must make guesses about the reasons for the crime.  Importantly, a perceived bias victimization may still have an impact on both the victim and the community regardless of the veracity of the belief—perceptions need not match reality to influence behavior.

Based on this, the present work has two goals.  First, we seek to replicate the general question from the small number of prior studies (Lyons 2007, Gladfelter et al. 2017) on the ecological distribution of anti-white hate crimes in a new context: the majority-white city of Seattle.  Second, in addition to police reports of anti-white bias incidents, we also use self-report data from a large survey of residents of Seattle neighborhoods.  This allows us to ask a different kind of question, examining the neighborhood distribution of what we believe may be two different phenomena: incidents that are reported to the police specifically as something motivated by hate, and the theoretically larger set of crimes that are more simply believed by the victim to have been motivated by racial bias. 

These questions are important for several reasons.  First, anti-white hate crimes are particularly interesting because they do not clearly fit within the threat-based theoretical framework generally employed to make sense of other kinds of hate crimes—that hate crimes are perpetrated by majority or dominant group members against minority or subordinate group members whom they perceive as a threat to their privileges.  This question in particularly important in light of the recent U.S. political climate, which has included both a surge in hate incidents and hate groups (Southern Poverty Law Center 2016; Potok 2017) as well as claims of anti-white racism (Norton and Sommers 2011), often in response to those seeking racial justice or equality for non-white racial groups (e.g. Kew 2016).

Second, this work heeds the call to pay more attention to overlooked sources of bias in crime and justice processes (Peterson, 2017).  Typically, hate crime research focuses on the motivation of the offender.  However, the reality is that it is often the victim’s perception of the crime and offender that will determine whether the crime is considered motivated by hate or not.  Implicit racial biases are pervasive (e.g. Banaji and Greenwald 2013), and impact not just criminal justice processing (e.g. Banks et al. 2006; Pizzi et al. 2005), but also perceptions of crime more broadly (Eberhardt et al. 2004; Drakulich 2015; Drakulich and Siller 2015).  In this light, implicit bias likely plays a role not just in the way that offenders see victims, but also in the way that victims see offenders.  While much attention has been paid to bias on the part of the offender as the motivation for perpetrating a hate crime, little to no work has directly considered potential bias—implicit or explicit—in the way victims see the offender. Though we do not examine implicit bias directly, we draw on this body of work to develop expectations about the way individuals may perceive a victimization, and in particular how those perceptions may be influenced by the racial context of the neighborhood.

THEORETICAL MODEL 

Prior work has drawn on some version of three different theories to set up explanations for the distribution of hate crimes across neighborhoods (e.g. Lyons 2007). First, resource threat theories position hate crime as the end result of perceived competition for scarce economic and political resources. From this perspective, residents perceive an incoming marginalized out-group population as a threat to valued but limited economic opportunities (Brief et al. 2005; D’Alessio, Stolzenberg, and Eitle 2002). This perceived threat leads to prejudice, discrimination, and negative stereotyping of the out-group (Brief et al. 2005; Jackson 1993; Schofield 2009; Stephan and Stephan 2000).

By contrast, the defended neighborhood thesis argues that the motivation for unrest is not based on competition for economic resources but rather on cultural competition. An influx of in-migration by a racial or ethnic minority may challenge the shared cultural identity of the neighborhood (Blalock 1967; Bobo 1988; DeSena 2005; Grattet 2009; Green, Strolovitch, and Wong 1998; King, Messner, and Baller 2009; Levin and McDevitt 2002; Suttles 1972). Under some conditions, the response to this perceived cultural threat is to ‘defend’ the territory from the invasion of ethnic outsiders (Buell 1980). Such defenses might include hostile acts of intimidation aimed at the offending minority group, such as damaging property belonging to the group or physically assaulting a member of the group. Lyons (2007) argues that these defenses are more likely in neighborhoods which are economically and socially organized, as these areas have the resources available to actively repel these perceived threats. Green and colleagues (1998) found support for this when examining racial change over a ten-year period, identifying that hate crime was highest when in-migration occurred in areas where the majority of the population was white. They identified that it is not simply the proportion of a given minority in an area, but rather the change in this proportion that leads to conflict. This led the authors to suggest that hate crime “stems not from economic frustration but from an exclusionary impulse on the part of whites defending what they perceive to be their territory” (Green, Strolovitch, and Wong 1998:373).

Finally, scholars have drawn on a more general theory of crime: social disorganization. According to this theory, high levels of residential instability, ethnic diversity and economic disadvantage decrease social cohesion and the willingness of residents to engage in informal social control (Sampson, Raudenbush, and Earls 1997; Shaw and McKay 1942). The result of this breakdown of social relationships is increased crime rates, which will include both hate and non-hate offences (Brunton-Smith, Jackson, and Sutherland 2014; Hirschfield and Bowers 1997; Sampson et al. 1997; Villarreal and Silva 2006). From this perspective, crime and hate crime would share the same ecological drivers given that they both involve a criminal element (Green et al. 1998; Lyons 2007). Grattet (2009:135) argues that “bias crime always contains conduct that is already criminalized,” meaning that hate crimes are already reflected in counts of non-hate crimes, for instance as vandalism, a threat, or an assault.  

However, there are problems in applying each of these theories to our specific problem of interest: anti-white hate crimes. First, both the resource threat and defended neighborhoods perspectives are rooted in a notion of threat. Classically, racial threat theories have focused on threats posed by minority or subordinate group members as perceived by majority or dominant group members (Blalock 1967). This is rooted in a sociological notion of prejudice as a sense of group position, in which the perceived threat emerges from a feeling of group-based proprietary claims to certain privileges or advantages combined with the feeling that a subordinate group is interested in obtaining those privileges or advantages (Blumer 1958). In Seattle, as in the United States in general, white residents clearly occupy the dominant socio-political and economic racial position. Broadly, then, theories rooted in this notion of perceived subordinate group threat will be less useful for explaining hate crimes against dominant groups.  Confirming this, Lyons (2007) fails to find evidence to support either of these explanations as a predictor of anti-white hate crimes.  However, on a smaller scale, when a process like gentrification increases the presence of dominant group members in an area with an established non-dominant group population, this may be experienced as a threat.   

The third theory—social disorganization—presents its own issues. The most important implication of this perspective is that it assumes that hate crimes do not require a special explanation—that they are etiologically indistinct from other more routine kinds of crimes (e.g. Lyons 2007). The issue, then, is that there is no theoretical reason—at least from social disorganization theory—to expect that this model would help explain hate crimes against some groups but not others. However, Lyons’s (2007) work found just this to be the case (as did Gladfelter et al. 2017).  Thus, while anti-black hate crimes may be best understood as specialist crimes—motivated primarily by the message they send to the group perceived as posing a threat—anti-white crimes may, at least in part, be rooted in the same structural causes of many non-hate crimes, suggesting they may be a fundamentally different phenomenon than hate crimes against subordinate groups (e.g. Lyons 2007).  Police reports of anti-white hate crimes, then, may be most likely in the kinds of places where police reports of non-hate crimes are also common. 

These cases reported to the police, however, represent only a small portion of a larger population of incidents believed by the victim to have been motivated by racial bias or animus.  This larger group of incidents—likely varying widely in the seriousness of the offence and the explicitness of the racial motivation—may require a separate explanation.  The starting place for the explanation is the idea that the motivation behind a crime can be difficult to ascertain with any certainty. In fact, these issues may be more pronounced for anti-white hate crimes, as they tend to occur in contexts where larger numbers of non-hate crimes also occur, and may be less likely to be committed by hate-crime specialists (Lyons 2007).i As a result, white victims may be more apt to make perceptual misjudgments.

After experiencing a victimization, the victim is left with a difficult question: why? Narratives which ascribe meaning, attribute blame, and suggest future actions are important (see, for instance, sociological work on narratives and framing: Goffman 1974; Benford and Snow 2000; Small, Harding, and Lamont 2010). Symbolic interactionism suggests that “human beings act towards things on the basis of the meaning that the things have for them” (Blumer 1969 p. 2), but that the meaning of a thing is not necessarily intrinsic to that thing. In short, narratives and meaning may be important independent of the real world events and conditions they interpret. Applied to crime, prior work has suggested victims hold differing views of the blame for and motivation behind their victimizations (e.g. Perilloux, Duntley, and Buss 2014; Lim, Valdez, and Lilly 2015). The question then, is how people come to understand their victimization experiences, especially when the motivations or causes are ambiguous. To answer this, two insights about our cognitive processes are useful. 

First, our views and judgements are often influenced by subtle cues in an automatic fashion without necessarily consciously realizing that this has happened, a phenomenon known as priming (e.g. Kahneman 2011; Banaji and Greenwald 2013). Race is a powerful prime (e.g. Banaji and Greenwald 2013). This is particularly true for perceptions of crime.  The racial context—and in particular the presence of larger numbers of African-Americans—provokes overestimations of the presence of disorder, the threat of criminal danger, and the likelihood of victimization (e.g. Quillian and Pager 2001, 2010; Sampson 2009; Pickett et al. 2012; Drakulich 2012, 2013). If someone has been primed by the racial context, they may be more likely to think of race as they attempt to make sense of a victimization experience, and thus more likely to attribute a causal role to race. In trying to determine the reason for a victimization, white residents of black neighborhoods may be more likely to assume race must have played a role. In short, the racial context may influence not just estimations of the likelihood of victimization, but also understandings of the experience of victimization. 

Second, ascertaining whether a victimization was motivated by racial hate is, at its root, a judgement about probability in the face of uncertain or imperfect information. Social psychologists suggest that in these kinds of questions, we rely on heuristics—or cognitive tricks—to develop answers (Tversky and Kahneman 1974; Kahneman 2011). The problem is that these heuristics often lead us to the wrong answer, and do so in relatively systematic and predictable ways. For instance, if characteristics of the victimization are a good match for a stereotypical mental image of an anti-white hate crime—for example that they tend to occur in areas with more black residents—the victim may be more likely to identify it as such, even though in reality non-hate crimes may be more common than hate crimes in these same places.ii In short, perceptions that victimizations were motivated by anti-white bias may be the product of a systematic patterning of assumptions about the motives for victimizations primed by the racial context.

PRIOR RESEARCH    

ANTI-WHITE HATE CRIME

As suggested above, prior work has suggested that the distribution of police reports of anti-white hate crimes appear to be the product of social disorganization processes.  Lyons (2007) finds that anti-white incidents in Chicago were reported at a higher frequency within socially disorganized neighborhoods, particularly where there had recently been a high population turnover. In contrast, neighborhoods with higher levels of informal social control were associated with fewer anti-white hate crimes, suggesting that informal social control has a mediating effect on the number of hate crimes reported within the community. Gladfelter and colleagues (2017) found that in Pennsylvania, hate crime incidents were significantly and consistently associated with residential instability. The role of advantage, on the other hand, appeared to depend on race: anti-black crimes were most likely in largely white and socio-economically advantaged communities while anti-white hate crimes were most likely in socio-economically disadvantaged neighborhoods (Gladfelter et al. 2017).  Thus, in both studies, antiwhite hate crimes appear better explained by a classic social disorganization model (the confluence of residential instability and socio-economic disadvantage). 

These findings have an important implication for an understanding of hate crimes: the ecological drivers of hate crimes against different races may be distinct. Lyons (2007) finds that anti-black crimes were more likely to occur in organized, relatively affluent communities that had recently experienced a sudden influx in black in-migration. Organized communities, rich in economic resources, were able to work together to exhibit hate crime as a ‘message’ crime. Largely consistent with this, Gladfelter et al. (2017) report anti-black crimes are most likely in residentially unstable but affluent and racially homogenous communities.  In short, anti-white hate crimes appear most likely in socially disorganized neighborhoods—along with other crimes more generally.  Anti-black hate crimes, by contrast, appear more likely in relatively organized communities experiencing some kind of change (residential instability generally in Gladfelter et al.’s 2017 work and an influx of black residents specifically in Lyons’ 2007 work).  Consistent with this, Lyons (2008a) finds that “the influx of black newcomers matters only in white communities whose residents express a strong sense of community attachment” (p. 378, original emphasis).

PERCEIVED AND REPORTED HATE CRIME INCIDENTS

Studies on the neighborhood ecology of hate crime have to date primarily relied on official data to inform their analyses (for example Green et al 1998; Grattet 2009; Lyons 2007).iii However, there are several problems noted with official data in relation to hate crime offences.  First, police records in many jurisdictions are inaccurate or incomplete (Iganski 2001; Levin 1999; Perry 2001). In the United States, hate crime data should be reported annually to the Federal Bureau of Investigations for the Uniform Crime Reports under the Hate Crime Statistics Act 1990 (Krouse 2010). However in 2006 almost 5,000 police departments out of approximately 17,000 departments in the nation did not provide figures (ADL 2007; Henry 2009). Further, 83.3% of the 12,620 agencies that did participate reported zero hate crimes in their jurisdiction. Errors have also been identified in the hate crime data collection, as often agencies that do not provide data will be reported as recording zero offences (Southern Poverty Law Center 2001). Police often do not understand whether to classify an offence as a hate crime or are reluctant to do so to avoid additional paperwork (Henry 2009).

Additionally, even when police agencies do record and report hate crimes, victims often do not report the crimes to the police.  Estimates suggest sixty to seventy percent of incidents are not reported – a rate which is reasonably consistent across the United States, United Kingdom and Australia (Bureau of Justice Statistics 2014; Home Office 2013; VEOHRC 2010). Victims may be reluctant to report victimization to police for a number of reasons, including fear of secondary victimization, mistrust of the police, unsatisfactory previous experiences with the police, or perceiving the offence to be too trivial to report (e.g. Harlow 2005). Additionally, victims may be less likely to report the crime to the police—or to report it as a normal non-hate crime—in cases where the evidence of the hate is less clear. In general, there appears to be variability in what people perceive to be a hate crime (e.g. Craig and Waldo 1996).  The potential detrimental effects of hate crimes are not limited to those which end up reported to the police. For this reason, we are interested not just in those hate crimes which are reported to the police, but also in the full population of crimes perceived to be motivated by hate, regardless of whether they were reported to the police.

THE PRESENT RESEARCH

Anti-white hate crimes appear to be relatively prevalent. According to the FBI (2014), 21.4 percent of the 3,407 racially motivated hate offenses reported in the UCR in 2013 were motivated by an anti-white bias. Despite this, as Perry (2002, 2003) has noted, scholars of hate crime research have tended to neglect the dynamics of anti-white victimization.

In this research, we evaluate the neighborhood etiology of hate crime offenses using both self-reported victimization data from the Seattle Neighborhoods and Crime Survey and hate crimes reported to the Seattle Police Department. We are interested in exploring several basic research questions. The first is to investigate the neighborhood conditions in which anti-white-crimes are most prevalent. The theoretical models outlined above suggest two possible explanations. Based on Lyons’s (2007) work, our first expectation is that anti-white bias crimes are most prevalent in classically socially disorganized communities: those with disadvantaged structural conditions, low collective efficacy, and high general crime rates. The second possibility involves the notion of racial assumptions or misperceptions. Under this explanation, white residents may be more likely to interpret victimizations as racially motivated when they live near larger numbers of black neighbors. In other words, the racial context may influence the meaning people attribute to victimization experiences, just as prior work has found the racial composition to influence perceptions of crime more generally (e.g. Quillian and Pager 2001; Drakulich and Siller 2015). 

The second question contrasts our two measures of hate crimes, asking whether there is a difference between a general measure of self-reported hate crimes and those reported to the police. It is often difficult to establish hate as a motive for crimes, and it may be that citizens frequently experience victimizations which they interpreted as motivated by hate, but for which not enough direct evidence of that hate exists to make it worthwhile to file a formal police report. Based on Lyons’s (2007) findings, we expect that police reports of anti-white hate crimes will be more likely in socially disorganized neighborhoods. Perceived bias victimizations may also be more common in socially disorganized neighborhoods. However, to the degree that these self-reported perceptions reflect a wider range of incidents, including more cases for which the evidence of bias is less concrete, assumptions and misperceptions may play a larger role. In particular, we expect that the racial context of the neighborhood may play a larger role in the distribution of perceptions of anti-white hate incidents.

DATA, MEASURES & METHODS

To test these ideas, we draw on several sources of data from Seattle, Washington collected during the early 2000s. Seattle presents an interesting setting in which to examine hate crimes targeting whites. Relative to cities in the Midwest and Northeast, Seattle is simultaneously relatively white and relatively integrated (Crutchfield, Matsueda & Drakulich, 2006). Significantly for this topic, during this period of time, Seattle was undergoing rapid gentrification, specifically middle and upper class white residents moving into neighborhoods that were traditionally less white (Crutchfield et al, 2006; Fay-Ramirez, 2015).

DATA

The first source of data is the Seattle Neighborhood Crime Study (SNCS), a survey of 5,812 residents of 123 Seattle census tracts conducted in 2002 and early 2003, with an AAPOR cooperation rate of 97 percent and a response rate of 51 percent.iv Four separate sampling schemes were employed in data collection. These include two simple random samples, the first drawn from the white pages and the second drawn from the universe of residential addresses without listed numbers. A third sample was designed to be comparable to an older survey of Seattle residents, while a fourth sample disproportionately targeted blocks with the largest proportions of ethnic and racial minorities (Matsueda 2010; see Drakulich 2013 for more details). The survey provides our key perceptual measure of anti-white hate incidents as well as our measure of neighborhood collective efficacy. 

A second source of data on bias victimizations is developed from Seattle Police Department (SPD) reports on bias crimes. The data come from a public disclosure request for “bias crimes incidents” reports compiled by the Seattle Police Department from 2000 to 2005. The request was made by Ken Molsberry, who eventually produced a 2006 report in collaboration with the Seattle LGBT Community Center. In addition to the report, Ken Molsberry made scans of the paper copies of the bias crime incident reports publicly available (specific addresses and other potentially identifying information were redacted by the SPD). Using this information, we developed an incident-level data-set noting key characteristics of the incident as well as the census tract in which the location occurred.v

Information on the local context was constructed from two additional sources of data: socio-demographic and economic information from the 2000 U.S. Census and information on the prevalence of general crimes (not just bias incidents) from the Seattle Police Department for each census tract targeted by the survey.

MEASURES

Perceived Bias Victimizations. The core interest of the paper is victimizations perceived by white respondents to have been motivated by race-ethnic bias. The SNCS includes a variety of victimization reporting questions. After asking respondents to report on three kinds of personal victimizations—vandalism to one’s home, verbal threats, and physical assaults—respondents were asked whether “any of the incidents just mentioned, involving vandalism, harassment, or assault, [were] motivated by dislike for people because of their race or ethnic background.” Respondents were asked to report all incidents which have occurred since childhood, but the results presented here are limited to incidents occurring in the last two years.vi Though these offences range in seriousness, they are all examples of police-reportable offences.

Table 1 presents basic counts and rates of perceived bias victimizations by the race and ethnicity of the respondent. Given the larger number of white respondents—representative of the majority-white population of the city of Seattle—the largest number of victimizations was unsurprisingly reported by non-Hispanic whites. At the same time, however, non-Hispanic white respondents reported such incident at a lower rate than any other race-ethnic group. Asian-Americans also reported relatively low rates, with Hispanic, African-American, and “other” race-ethnicities (including native Americans) reporting much higher rates of perceived bias victimizations. In the analysis we focus specifically on victimizations perceived to be motivated by racial bias and reported by non-Hispanic white respondents. 

***TABLE 1 ABOUT HERE***

Police Reports of Bias Incidents. We also construct a measure of bias incidents from the SPD “bias crimes incidents” reports. We restrict the measure to capture incidents reported to the police which have a “bias focus” on race (as opposed to sexual orientation or religion, for instance), and for which the victim(s) were reported to be non-Hispanic white. The resulting measure is a count of such incidents for each census tract. Consistent with our claim that the police reports reflect a selective sub-class of perceived bias incidents, the rate of such victimizations is far lower than that reflected in the victimization data: only about 1 in every 10,000 white Seattle residents were the victim of bias incidents between 2000 and 2005 according to the police reports.

Neighborhood Racial Composition, Socio-Economics, and Crime. Several measures of neighborhood context were developed from the 2000 US Census. The racial composition is captured as the proportions Asian, African-American, and Hispanic. Increases in the white population are captured as the difference between the proportion non-Hispanic white in 2000 and 1990.  Concentrated disadvantage is captured through four indicators: the proportion of families in poverty, the proportion on public assistance, the proportion unemployed, and the proportion single-parent households.vii Residential stability is captured by two indicators: the percentage of residents who lived in the same home for at least five years and the proportion of home owners. Additionally, an indicator of the general presence of all types of violent crimes (rather than just bias incidents) is included. The measure reflects violent crimes known to the Seattle Police Department—captured as the average number of crimes per year per thousand residents, averaged over a three-year period from 1999 to 2001.

Collective Efficacy. To capture a neighborhood social process potentially relevant to the distribution of bias incidents we include a neighborhood-level measure constructed from the survey data. Collective efficacy is captured by six indicators. Two of these tap into perceived trust and cohesion in the community, and four ask about perceptions of neighbor’s willingness to intervene when youths are misbehaving.viii Neighborhood-level estimates of collective efficacy were captured as empirical Bayes residual scores from multi-level models controlling for basic personal characteristics that may be associated with bias in the reporting of neighborhood social processes.ix In short: neighborhood-level estimates of the presence of collective efficacy were constructed by treating respondents as reporters of its presence, controlling for the potentially uneven distribution of socio-demographic characteristics that may be associated with biases in this reporting, and then adjusted based on the confidence of particular neighborhoods as a function of the number of reporters within that neighborhood. 

ANALYTIC STRATEGIES

The two bias incident outcomes are captured at the neighborhood-level as counts. The classic model for count data is the Poisson, though this is frequently not the best-fitting model either because of over-dispersion (recommending alternatives like the negative binomial model) or because of an excess of zeroes (recommending a zero-inflated model). However, the counts of bias crimes did not show evidence of either of these problems,x and as such Poisson models are presented here. In all cases we report robust standard errors and corresponding tests of significance. For both models, the white population is included as an exposure term. The models of police reports include the logged population of white respondents per neighborhood from the census as a control, while the victimization reports include the logged number of white survey respondents.  We also ran all of the models using the total population rather than just the white population as the exposure term and found substantively identical results. 

LIMITATIONS AND METHODOLOGICAL CONSIDERATIONS

Fortunately, race bias incidents, particularly those reported by whites, are relatively rare. This rarity, however, does pose challenges in estimating the neighborhood prevalence of such incidents using the survey data. A combination of the low incidence and the ordering of questions in the SNCS also prevents us from separating out types of victimization that involve bias (for instance distinguishing harassment from assault). However, our substantive interest is less in the event itself and more in the perception that the event was motivated by race-ethnic bias. The coding of the police reports is consistent with this interest. The survey also does not include any information about the offender. The low incidence of cases for non-white Seattle residents, combined with the geographic concentration of non-white Seattle residents in a small number of neighborhoods, means that neighborhood-level estimates of non-white hate crime victimizations are not reliable.xi  Thus, a direct comparison study of neighborhood variation in non-white hate crime victimization is not possible (although we recommend future work explore direct comparisons of white and non-white perceptions of bias incidents).  Finally, although the low incidence is limiting, there are substantial generalizability advantages to the random sampling design relative to a more purposive design that may have netted a larger number of bias crime incidents.

RESULTS

The results are organized in two sections. The first presents bivariate associations between the two measures of anti-white bias incidents and measures of the neighborhood context, examining whether basic similarities or differences exist in the distribution of these incidents over Seattle neighborhoods. The neighborhood characteristics examined are highly interdependent: a combination of segregation and inequality produces substantial overlap between the race-ethnic composition and the socio-economic structural conditions, and collective efficacy and crime are often described as products of these structural factors. Given this, the second section examines these characteristics in a multivariate context, simultaneously exploring the potential roles of social disorganization and the racial composition.

BIVARIATE ASSOCIATIONS

Table 2 presents bivariate associations between the two measures of anti-white bias incidents and measures of neighborhood characteristics.xii These incidents appear to occur most often in neighborhoods characterized by classically socially disorganized structural conditions (high disadvantage, low stability, and high ethnic heterogeneityxiii) as well as an absence of social processes often associated with lower levels of crime overall (collective efficacy). This is true regardless of which measure of bias incidents is used, and is generally consistent with the finding in prior work (Lyons 2007) that anti-white bias incidents—unlike bias incidents targeting other groups—are the product of the same kinds of processes that produce non-bias victimizations.xiv Consistent with this story, such incidents were also more common in neighborhoods with higher overall levels of violent crime.

***TABLE 2 ABOUT HERE***

The major difference between the two measures appears to be related to race. The perceived bias victimizations—but not the police reports of bias incidents—are substantially more likely in neighborhoods with larger African-American populations. Notably, this means that while police reports of anti-white bias incidents do not appear more likely—either in Seattle or in Chicago in Lyons’s work (2007)—in neighborhoods with larger African-American populations, individual perceptions of bias incidents, many of which may have gone unreported to the police, appear to be highly more likely in such neighborhoods.

Finally, police reports of anti-white bias incidents are not more likely in neighborhoods having experienced an increase in the proportion white either in Seattle or in Lyons’s (2007: 845-846) work on Chicago.  Thus, in both places, a “white threat” hypothesis appears to be a poor explanation for anti-white bias incidents—at least those reported to the police.  Perceived bias victimizations, however, do appear more common in neighborhoods with growing white populations.  Table 3 explores this finding further.  The first column presents correlations between an increase in the white population and the racial composition of the community.  The results suggest that the white population was less likely to increase in communities with larger numbers of Asian or Hispanic residents, and not any more or less likely in communities with larger existing white populations.  Instead, increases in the white population were most likely in neighborhoods with larger black populations.  Given this overlap, the second column asks whether perceived anti-white bias crimes are more related to the proportion black or increases in the white population when considered simultaneously.   The proportion black remains strongly and positively associated with perceptions of anti-white bias victimizations.  By contrast, increases in the white population are no longer associated with perceived bias victimizations when the racial composition is controlled for.  In short, it is the presence of black residents and not increases in white residents that seems to be driving perceptions of bias victimizations among white respondents. 

***TABLE 3 ABOUT HERE***

NEIGHBORHOOD MODELS OF ANTI-WHITE INCIDENTS

Of course, the neighborhood characteristics described in Table 2 are highly interdependent: neighborhoods with larger number of racial and ethnic minorities tend to have worse structural conditions and higher crime rates. Thus, our second question is which of these neighborhood characteristics appears most important to the distribution of bias incidents when considered simultaneously. Table 4 presents three models for each of the two measures of anti-white bias incidents.xv For each outcome, the first column presents the basic social disorganization model: structural characteristics and collective efficacy, the second adds violent crime, while the third adds measures of the racial composition.

***TABLE 4 ABOUT HERE***

Police reports of incidents were less likely to occur in neighborhoods characterized by residential stability (Model 1 of Table 4). Disadvantage, heterogeneity, and collective efficacy are not directly related to the distribution of police-reported anti-white incidents once stability is accounted for.  Police reports of incidents were more likely to occur in neighborhoods with a higher overall violent crime rate (Model 2).  Interestingly, violent crime exerts an independent influence, and does not explain the connection between residential instability and police reports of bias incidents.  Model 3 confirms the story from the bivariate results that the racial composition does not appear relevant to the distribution of anti-white bias incidents known to the police.

For self-reported perceptions that victimizations were motivated by bias, a somewhat different pattern emerges. When considering the classic social disorganization variables, perceived bias incidents appear somewhat less likely in neighborhoods with higher levels of collective efficacy (Model 4 of Table 4) and somewhat more likely in neighborhoods with high violent crimes rates (Model 5). Both of these effects, however, disappear when accounting for the role of the racial composition (Model 6).xvi Perceptions that victimizations were motivated by bias were substantially more likely in neighborhoods with a higher proportion African-American. This effect is relatively strong. Most Seattle neighborhoods have relatively low numbers of African-Americans. For those with essentially no African-Americans, the predicted number of perceived bias incidents would be less than one (.60); similarly, for the average neighborhood which is only about 8 percent African-American, the predicted number of perceived anti-white bias incidents would also be less than one (.84). However, in a neighborhood that is one-quarter black, the predicted number of incidents increases to more than one and a half (1.67) and in a neighborhood which is half black (the highest percentage in Seattle), the predicted number of incidents is nearly 5 (4.67). 

DISCUSSION

The goal of the present work was to address two questions. The first concerns the contextual factors associated with the distribution of anti-white hate incidents over neighborhoods. The second involves potential differences between those incidents reported to the police and the larger population of incidents simply perceived by the victim to have been motivated by race. The results appear to suggest that different factors explain the distribution of perceived and reported bias incidents. Anti-white hate incidents reported to the police are more common in neighborhoods in which other kinds of street crimes are also likely, and may be committed by those who also commit other kinds of crimes. In short, they may not be strongly distinguished in their origins from other kinds of crime. On the other hand, white people who simply believe that they have been a victim of a hate crime were most frequently found in neighborhoods with larger numbers of black residents.

The first finding is consistent with the pattern identified by Lyons (2007): anti-white bias incidents appear to follow roughly the same pattern as other kinds of neighborhood crime not motivated by bias, communities characterized by low residential stability and high levels of other kinds of crime—and, notably, such incidents are not more common in places where the white population has been increasing. This is significant because it differs from the spatial patterns Lyons (2007) identifies for anti-black bias crimes—Gladfelter et al. (2017) also report this distinction. Lyons (2007) suggests that in contrast to anti-black hate crime offenders whose crimes serve a specialized purpose to intimidate and defend based on racial prejudice, anti-white offenders may engage in a wide range of criminal activities with a number of different motivations which may or not include motivations of prejudice.

The second finding moves us beyond prior work, which has not previously explored the distribution of perceived anti-white bias crimes across neighborhoods. The results, interestingly, suggest a distinct pattern from those which are reported to the police. Specifically, the primary neighborhood determinant for perceived anti-white hate incidents is the racial composition of the neighborhood, and in particular a higher proportion African-American.  In other words, the story about where white residents are most likely to perceive that incidents are motivated by hate is simple: they do so in areas with more black neighbors. By contrast, hate incidents which are reported to the police do not appear to be any more likely in neighborhoods with larger black population, even at the bivariate level.xvii

Our interpretation is that the police reports likely reflect a subset of cases for which the evidence of the hate motivation is more overt, while the perceived incidents capture a much wider array of victimization incidents, many of which may be more ambiguous in their motivation. This greater ambiguity allows the perceptual biases to play a larger role.  In these cases, then, the racial context may be acting as a prime and providing an important interpretive lens through which white residents perceive their experiences and make judgments about the motivation of others.

CONCLUSION

This work advances our understandings of hate crimes by exploring two understudied dimensions of this issue which reveal important differences from existing work. First, the vast majority of prior work on hate crimes has focused both theoretically and empirically on hate crimes directed at members of minority or lower-status groups. Despite this, large numbers of white Americans report being the victim of bias incidents motivated by their race. In fact, in majority white cities like Seattle, these may represent the majority of racial hate incidents, even if whites are victimized at lower rates. In addition to being understudied, anti-white hate incidents—those perpetrated against members of the dominant majority group—also present an interesting counter-example to the kinds of hate incidents usually theorized and studied. In the present work we find that in fact anti-white hate incidents appear to have a fundamentally different etiology than prior work has suggested for those perpetrated against other groups. 

Second, the vast majority of prior work on the neighborhood distribution of hate crimes has been limited to those incidents reported to the police. This misses a larger and potentially distinct subset of incidents that do not end up reported to the police. This omission appears consequential, as our work finds a fundamentally different distribution of perceived versus reported hate crimes, with each more common in neighborhoods with different social and demographic characteristics. 

These findings have several important implications for theory and suggestions for future research. First, confirming work by Lyons (2007) and Gladfelter et al. (2017), anti-white hate incidents appear to be fundamentally distinct in their etiology from anti-black incidents. While Lyons (2007) found that anti-black incidents appear consistent with a racial threat model—organized communities experiencing an increase in black residents—anti-white incidents appear more common in the kinds of socially-disorganized communities that experience larger numbers of non-hate crimes. If anti-white and anti-black hate crimes are the product of fundamentally different theoretical processes and appear in different kinds of neighborhoods, this suggests we may need to rethink whether they are fundamentally different phenomena—just as scholars suggest there are fundamental differences between “racism” and “reverse racism.” 

This finding also suggests several important new lines of research. First, if these incidents are motivated by hate or dislike of whites as a group, what is the source of this hate? For incidents targeting non-whites, threat-based models have been used to understand the genesis of animus toward members of minority or subordinate groups, but these make less theoretical sense for incidents targeting members of majority or dominant groups. Social disorganization—the theory which seems to best explain the distribution of anti-white incidents reported to the police—does not suggest any obvious mechanisms by which anti-white hate would be produced. If the hate is the product of different theoretical processes, this may reinforce the above suggestion that anti-white hate incidents be treated as a fundamentally different phenomenon.

Second, future work should also explore the suggestion that white residents may be subject to greater errors in judging whether incidents are motivated by anti-white hate, in part due to their greater co-occurrence with non-hate crimes.  If this is the case, it suggests a need for better understanding the cognitive and perceptual processes by which people make determinations about the causes of their victimizations. 

Finally, this work adds a new dimension to our understanding of how the neighborhood racial context shapes perceptions of crime. Previous work has found that residents of neighborhoods with larger numbers of black neighbors overestimate the presence of disorder, the threat of criminal danger, and the likelihood of victimization (e.g. Quillian and Pager 2001, 2010; Sampson 2009; Pickett et al. 2012; Drakulich 2012, 2013). The present work finds that the presence of black neighbors may make white residents more likely to assume that victimizations were motivated by race. In other words, the racial context may shape not just our view of the danger posed by crime, but also how we interpret the direct experience of crime. 

Prior work has suggested that racialized misperceptions of danger have consequences: residents may flee or withdraw, they may avoid participating in formal or informal social control efforts, all of which harm those communities (e.g. Drakulich and Crutchfield 2013; Miles-Johnson et al. 2016). Perceptions that victimizations were motivated by anti-white hate may have similarly negative consequences, including flight, withdrawal, or the avoidance of inter-racial contact, ironically one of the few mechanisms that may reduce racial misunderstandings (Allport 1954; Drakulich 2012).  Future work would do well to explore these possibilities. 

This basic idea—that the racial context primes not only how we view our local space but how we interpret specific experiences within that space—also suggests a range of other investigations. How, for instance, does the racial context shape our interpretation of a neighborhood asking for a favor, a needy person asking for change, or an interaction with a law enforcement officer. In any case, this works points toward the pervasive, complicated, and multifaceted ways in which the racial context influences neighborhood life. 

REFERENCES

Allport, Gordon W. 1954. The Nature of Prejudice. New York: Doubleday Books.

American Association for Public Opinion Research. 2011. Standard Definitions: Final Dispositions of Case Codes and Outcome Rates for Surveys. 7th edition. AAPOR.

Banaji, Mahzarin R., & Anthony Greenwald. 2013. Blindspot: The hidden biases of good people. New York: Random House.

Banks, R. Richard, Jennifer L. Eberhardt, and Lee Ross.  2006. “Discrimination and Implicit Bias in a Racially Unequal Society.” California Law Review 94:1169-1190.

Benford, Robert D., and David A. Snow. 2000. "Framing Processes and Social Movements: An Overview and Assessment." Annual Review of Sociology 26: 611-639.

Benier, Kathryn and Rebecca Wickes. 2015. “The effect of ethnic diversity on collective efficacy in Australia.” Journal of Sociology 52: 865-73

Benier, Kathryn, Rebecca Wickes, and Angela Higginson. 2015. “Ethnic Hate Crime in Australia: Diversity and Change in the Neighbourhood Context.” British Journal of Criminology 56: 479-96.

Blalock, Hubert. 1967. Toward a theory of minority group relations, New York: John Wiley.

Blumer, Herbert. 1958. “Race Prejudice as a Sense of Group Position.” Pacific Sociological Review. 1:3-7.

Bobo, Lawrence. 1988. "Attitudes Toward the Black Political Movement: Trends, Meaning, and Effects on Racial Policy Preferences." Social Psychology Quarterly 51: 287-302.

Brief, Arthur P., Elizabeth E. Umphress, Joerg Dietz, John W. Burrows, Rebecca M. Butz, and Lotte Scholten.  2005. “Community matters: Realistic group conflict theory and the impact of diversity.” Academy of Management Journal48(5), pp.830-844.

Brunton-Smith I. Jackson J. and Sutherland A. 2014. “Bridging Structure and Perception: On the Neighbourhood Ecology of Beliefs and Worries About Violent Crime.” British Journal of Criminology 54: 503-26.

Buell Jr, Emmett H. "Busing and the defended neighborhood South Boston, 1974-1977." Urban Affairs Quarterly 16.2 (1980): 161-188.

Bureau of Justice Statistics. 2014. “Hate Crime Victimization, 2004–2012—Statistical Tables.” Available online at http://www.bjs.gov/index.cfm?ty=pbdetail&iid=4905

Craig, Kellina M. and Craig R. Waldo. 1996. “So, what's a hate crime anyway?” young adults' perceptions of hate crimes, victims, and perpetrators.” Law and Human Behavior 20:113-29.

Crutchfield, Robert D., Ross L. Matsueda, and Kevin M. Drakulich. 2006.  "Race, Labor Markets, and Neighborhood Violence."  Pp. 199-220 in The Many Colors of Crime: Inequalities of Race, Ethnicity, and Crime in America, edited by Ruth D. Peterson, Lauren J. Krivo, and John Hagan.  New York: New York University Press.

D'Alessio, Stewart J., Lisa Stolzenberg, and David Eitle. 2002. "The effect of racial threat on interracial and intraracial crimes." Social Science Research 31: 392-408.

DeSena J. N. 2005. Protecting One’s Turf: Social Strategies for Maintaining Urban Neighborhoods. United Press of America.

Drakulich, Kevin M. 2012. “Strangers, Neighbors, and Race: A Contact Model of Stereotypes and Racial Anxieties about Crime.” Race and Justice 2:322-55. 

Drakulich, Kevin M. 2013. “Perceptions of the Local Danger Posed by Crime: Race, Disorder, Informal Control, and the Police.” Social Science Research 42: 611-32. 

Drakulich, Kevin M. 2015. “Explicit and Hidden Racial Bias in the Framing of Social Problems.” Social Problems 62(3): 391-418.

Drakulich, Kevin M. and Laura Siller. 2015. “Presumed Danger: Race, Bias, Stigma, and Perceptions of Crime and Criminals.” Pp. 23-58 in Deadly Injustice: Race, Criminal Justice, and the Death of Trayvon Martin, edited by Devon Johnson, Patricia Warren, and Amy Farrell.  New York: New York University Press.

Drakulich, Kevin M. and Robert D. Crutchfield. 2013. “The Role of Perceptions of the Police in Informal Social Control: Implications for the Racial Stratification of Crime and Control.” Social Problems 60: 383-407. 

Fay-Ramirez, Suzanna. 2015. "The comparative context of collective efficacy: Understanding neighbourhood disorganisation and willingness to intervene in Seattle and Brisbane." Australian & New Zealand Journal of Criminology 48: 513-542.

Federal Bureau of Investigation. 2013. “Hate Crime—Overview.” Available online at http://www.fbi.gov/about-us/investigate/civilrights/hate_crimes/overview

Finn, Peter. 1988. “Bias Crime: Difficult to Define, Difficult to Prosecute.” Criminal Justice 3: 19-25 .

Gladfelter, Andrew S., Brendan Lantz, and R. Barry Ruback. 2017. “The Complexity of Hate Crime and Bias Activity: Variation across Contexts and Types of Bias.” Justice Quarterly 34: 55-83.

Goffman, Erving. 1974. Frame Analysis: An Essay on the Organization of Experience. Cambridge, MA: Harvard University Press

Grattet, Ryken. 2009. "The urban ecology of bias crime: A study of disorganized and defended neighborhoods." Social Problems 56: 132-150.

Green, Donald P., Dara Z. Strolovitch, and Janelle S. Wong. 1998. "Defended neighborhoods, integration, and racially motivated Crime 1." American Journal of Sociology 104: 372-403.

Green, Donald P., Laurence H. McFalls, and Jennifer K. Smith. 2001. “Hate Crime: An Emergent Research Agenda.” Annual Review of Sociology  , 27 : 479 – 504 .

Harlow, Caroline Wolf. 2005. Hate crime reported by victims and police. Washington, DC: US Department of Justice, Office of Justice Programs, Bureau of Justice Statistics.

Henry, Jessica S. 2009. “Hate Crime Laws: A Critical Assessment.” In Perry, B. and Levin, B. (eds.) Hate Crimes: Perspectives and Approaches Greenwood Publishing Group: Connecticut.

Hirschfield, Alexander, and Kate J. Bowers. 1997. "The effect of social cohesion on levels of recorded crime in disadvantaged areas." Urban Studies 34: 1275-1295.

Home Office. 2013. “An Overview of Hate Crime in England and Wales.” Available online at https://www.gov.uk/government/uploads/system/uploads/attachment_data/file/266358/hate-crime-2013.pdf

Iganski, Paul. 2001. "Hate crimes hurt more." American Behavioral Scientist 45: 626-638.

Jackson J. W. 1993. “Realistic Group Conflict Theory: A Review and Evaluation of the Theoretical and Empirical.” Psychological Record 43:395-413 .

Kahneman, Daniel. 2011. Thinking, fast and slow. New York: Farrar, Straus and Giroux.

Kew, Ben. 2016. “MILO: Black Lives Matter Is ‘The Last Socially Acceptable Hate Group’.” Breitbart.com. http://www.breitbart.com/milo/2016/12/13/milo-black-lives-matter-last-socially-acceptable-hate-group/

King, Ryan D., Steven F. Messner, and Robert D. Baller. 2009. "Contemporary hate crimes, law enforcement, and the legacy of racial violence." American Sociological Review 74: 291-315.

Krouse, W. J. 2010. Hate crime legislation. Washington, DC: Congressional Research Service. 
http://digitalcommons.ilr.cornell.edu/key_workplace/796

Lantz, Brendan, Andrew S. Gladfelter, and R. Barry Ruback. 2017. “Stereotypical Hate Crimes and Criminal Justice Processing: A Multi-Dataset Comparison of Bias Crime Arrest Patterns by Offender and Victim Race.” Justice Quarterly early online 13 Nov 2017.

Levin, Brian. 1999. "Hate crimes worse by definition." Journal of contemporary criminal justice 15: 6-21.

Levin, Jack and Jack McDevitt. 1993. Hate Crimes: The Rising Tide of Bigotry and Bloodshed. New York: Plenum Press.

Levin, Jack and Jack McDevitt. 2002. Hate Crimes Revisited: America’s War on Thos Who Are Different. Cambridge, MA: Westview Press.

Lim, Ban Hong, Christine E. Valdez, Michelle M. Lilly. 2015. “Making Meaning Out of Interpersonal Victimization: The Narratives of IPV Survivors.” Violence Against Women 21:1065-86.

Lyons, Christopher J. 2007. “Community (Dis)Organization and Racially Motivated Crime.” American Journal of Sociology 113: 815-63.

Lyons, Christopher J. 2008a. “Defending Turf: Racial Demographics and Hate Crime Against Blacks and Whites.” Social Forces 87: 357–385.

Lyons, Christopher J. 2008b. “Individual perceptions and the social construction of hate crimes: A factorial survey.” The Social Science Journal 45: 107-131.

Lyons, Christopher J. and Aki Roberts. 2014. “The Difference ‘Hate’ Makes in Clearing Crime: An Event History Analysis of Incident Characteristics.” Journal of Contemporary Criminal Justice 30:268-89. 

Matsueda, Ross L. 2010. Seattle Neighborhoods and Crime Survey, 2002-2003. ICPSR28701-v1. Ann Arbor, MI: Inter-university Consortium for Political and Social Research [distributor], 2010-12-10. doi:10.3886/ICPSR28701.v1

Miles-Johnson, Toby, Suzanna Fay-Ramirez, and Susann Wiedlitzka. 2016. “(Mis)Perceptions of Ethnic Group Size and Consequences for Community Expectations and Cooperation With Law Enforcement.” Race and Justice. Online ahead of print.

Molsberry, Ken. 2006. Bias Crimes and Incidents in Seattle: 2000 to 2005: An analysis by type of bias and neighborhood. Hate Crime Awareness Project, Seattle LGBT Community Center.

Norton, Michael I. and Samuel R. Sommers. 2011. “Whites see Racism as a Zero-Sum game that they are now losing.” Perspectives on Psychological Science, 6:215-18.

Perilloux, Carin, Joshua D. Duntley, and David M. Buss. 2014. “Blame attribution in sexual victimization.” Personality and Individual Differences 63: 81-86.

Perry, Barbara. 2001. In the Name of Hate: Understanding Hate Crimes. Routledge.

Perry, Barbara. 2002. “Defending the Color Line: Racially and Ethnically Motivated Hate Crime.” American Behavioral Scientist. 46: 72-92.

Perry, Barbara. 2003. “Where Do We Go From Here? Future Directions in Hate Crime Scholarship.” Internet Journal of Criminology.

Peterson, Ruth. 2017. “Interrogating Race, Crime and Justice in a Time of Unease and Racial Tension.” Criminology 55:245-272.

Pickett, Justin T., Ted Chiricos, Kristin M. Golden, and Marc Gertz. 2012. “Reconsidering the Relationship between Perceived Neighborhood Racial Composition and Whites’ Perceptions of Victimization Risk: Do Racial Stereotypes Matter?” Criminology 50:145-86.

Pizzi, William T., Irene V. Blair and Charles M. Judd. 2005. “Discrimination in Sentencing on the Basis of Afrocentric Features.” Michigan Journal of Race & Law 10: 327-353.

Potok, Mark. 2017. “The Year in Hate and Extremism.” Intelligence Report. Spring Issue.

Quillian, Lincoln, and Devah Pager. 2001. “Black Neighbors, Higher Crime? The Role of Racial Stereotypes in Evaluations of Neighborhood Crime.” American Journal of Sociology 107: 717-67.

Quillian, Lincoln, and Devah Pager. 2010. “Estimating Risk: Stereotype Amplification and the Perceived Risk of Criminal Victimization.” Social Psychology Quarterly 73:79-104.

Sampson, Robert J. 2009. “Disparity and Diversity in the Contemporary City: Social (Dis)order Revisited.” The British Journal of Sociology 60:1-31.

Sampson, Robert J., Stephen W. Raudenbush and Felton Earls. 1997. “Neighborhoods and Violent Crime: A Multilevel Study of Collective Efficacy.” Science 277: 918-24.

Schofield J. W. 2009. “Realistic Group Conflict Theory.” In Levine J. M. and Hogg M. A., eds., Encyclopedia of Group Processes & Intergroup Relations, SAGE Reference Online, 682-5.

Shaw, Clifford R. and Henry D. McKay. 1942. Juvenile Delinquency in Urban Areas. Chicago: University of Chicago Press.

Shively Michael. 2005. Study of Literature and Legislation on Hate Crime in America. National Institute for Justice. 

Small, Mario Luis David J. Harding and Michèle Lamont. 2010. “Reconsidering Culture and Poverty.” The Annals of the American Academy of Political and Social Science 629: 6-27.

Southern Poverty Law Center. 2016. Ten Days After: Harassment and Intimidation in the Aftermath of the Election. Southern Poverty Law Center Report. https://www.splcenter.org/sites/default/files/com_hate_incidents_report_final.pdf

Stephan W. G. and Stephan C. W. 2000. “An Integrated Threat Theory of Prejudice.” In Oskamp S., ed., Reducing Prejudice and Discrimination, 23-46. Lawrence Erlbaum Associates Inc.

Suttles G. D. 1972. The Social Construction of Communities. University of Chicago Press.

Tversky, Amos and Daniel Kahneman. 1974. “Judgment under Uncertainty: Heuristics and Biases.” Science 185(4157):1124-31.

Victorian Equal Opportunity and Human Rights Commission. 2011. “Third party reporting: What is it and why is it important?” http://www.humanrightscommission.vic.gov.au/media/k2/attachments/Reporting_Racism_Web_low_res.pdf

Villarreal, Andrés, and Braulio FA Silva. 2006. "Social cohesion, criminal victimization and perceived risk of crime in Brazilian neighborhoods." Social Forces 840: 1725-1753.

Notes

[i] Additionally, one of the key indicators used to identify race-motivated crimes is the use of racial slurs by the offender, either prior to or during the event.  However, many anti-white slurs are specific to rural or Southern whites. As a city in the northwest, these may hold less meaning for many white Seattle residents. 

[ii] Tversky and Kahneman (1974) refer to this heuristic as “representativeness” and the bias as an insensitivity to the prior probability of outcomes. 

[iii] A notable exception is Benier, Wickes, and Higginson’s (2015) study using survey data from Brisbane, Australia. 

[iv] The first number reflects the proportion who participated among eligible contacted units while the second reflects participation among all eligible units (American Association for Public Opinion Research 2011).

[v] The reports listed the census tract as identified by the reporting officer, but we found this frequently to be incorrect, so we re-coded the locations using the partial addresses available on the reports, which allowed the identification of the block. Additional clues from the reports plus a close look at Seattle maps allowed us to confidently identify the tract number in nearly all cases.  In the two cases that fell on the border between two tracts in a way that the tract could not be determined, the cases was randomly assigned to one of the two tracts.  Supplemental analyses showed that the exclusion of these cases had no impact on the substance of the results. 

[vi] Note: the answer categories available to respondents were yes, no, and maybe.  For the purposes of this analysis, only those responding “yes” were coded as having experienced a victimization perceived to be motivated by bias. 

[vii] In prior work from Chicago (e.g. Sampson et al. 1997), measures of concentrated disadvantage additionally include the proportion black, in part because of high levels of overlapping racial and class segregation.  Seattle is less segregated than Chicago, allowing the effects of race and economics to be considered separately (see Drakulich 2012; 2013; Drakulich and Crutchfield 2013). 

[viii] See Sampson et al. 1997 for a general description of the measure, and Drakulich et al. 2012 for a description of its construction from this survey. 

[ix] Empirical Bayes residuals “shrink” estimates toward zero by a factor related to their unreliability. The multi-level model includes person-level controls for age, sex, race-ethnicity, nativity, marital status, number of dependent children, years of education, household income, whether they own or rent their home, and the number of years they have lived in their home. 

[x] Tests for over-dispersion were not significant and the Poisson models produced an expected number of zeroes that was quite close to the observed zeroes in the counts of bias incidents (closer, in fact, than the expected number of zeroes from otherwise identical zero-inflated models). 

[xi] While all but 6 census tracts included at least 20 non-Hispanic white respondents from which to estimate the neighbourhood frequency of anti-white bias victimizations, in 104 of the 123 census tracts, fewer than 5 black respondents were interviewed. 

[xii] The table presents coefficients from a Poisson regression of incident counts from models.  Given the distribution of the outcomes, this is preferable to the presentation of linear correlations.      

[xiii] We include ethnic heterogeneity to be consistent with classical models of social disorganization (Shaw and McKay 1942; Benier and Wickes 2015) and to distinguish the effect of heterogeneity from the racial composition.

[xiv] In fact, these three findings—that anti-white bias incidents are more common in more disadvantaged, less stable, and lower informal social control neighborhoods precisely match the findings from Lyons’s (2007) final model (one that excludes some influential outliers). 

[xv] The bottom of table 4 presents several sets of diagnostics. The first line presents a basic test of overdispersion (that the estimate of dispersion is significantly greater than 1). The results—estimates close to one and a lack of significance—suggest that over-dispersion is not a problem. The second line examines the expected number of zeroes produced by the model.  In each case the expected number of zeroes closely matches the observed number of zeroes in the data.  This suggests zero-inflated models are not necessary.  Finally, the third line presents basic goodness-of-fit information.  The small residual deviance and lack of significance indicate a good fit.  AICs indicate a slight preference for the police report model with crime but without race and the perceived model with both crime and race. 

[xvi] Notably, the results show only modest evidence of multicollinearity, mainly affecting concentrated disadvantage and ethnic heterogeneity.  However, an inflated standard error does not explain the non-significant effect for concentrated disadvantage in any of the models, nor does it explain the lack of significance for ethnic heterogeneity in the first three models.  We do recommend caution in interpreting the role of ethnic heterogeneity in the sixth model (though notably the direction of the relationship is opposite that in the bivariate relationship). 

[xvii] Notably, Lyons (2007) also found the distribution of police reports of anti-white bias incidents to be unrelated to the percent black. 

TABLES

Comments
0
comment
No comments here
Why not start the discussion?